Wolfram Alpha: a shortlived fad

Not sure that matters. This is how these sorts of things work: Massive publicity at launch/open-beta, then a major drop-off as people realize "Oh, this app isn't the Jesus-app the PR made it seem like", then a gradual, permanent build-up as you slowly iterate and improve your site.

Furthermore, let's get rid of "Google-killer" as a metric. If Alpha can exist in a separate niche and be a useful service to people and make itself money, it's a win. It doesn't have to beat Google all-around to be useful to people or to make money.

We live in a ADHD society. Our 24-hour cable news and blog-centric Interwebz demands that things happen way faster than the real world works. The way people today make it sound, Wikipedia went from launch to "beating Britannica" in a few weeks, when in fact it took several years. Similarly, things like Wolfram Alpha are going to take years to go from launch PR as a "Google-Killer" to actually killing Google. The way our news system works, it expects news stories to unfold over hours, not months, and to be world-changing, not merely big.

It was too hard to use. I never knew how to phrase anything properly, so I just gave up on it.

Google trends is next to useless for this kind of thing. All I see is a ratio of the peak volume vs. a baseline "normal" traffic volume. Looking at quantcast and compete.. they went from zero to roughly 500k visitors a month in less than two months, plus they survived a massive traffic spike without a glitch. I call that a success in my books.

It's very handy, but specialist tool, almost like an almanac that computes useful facts in real time. Stuff I would normally use my Lee Valley pocket reference for (http://www.leevalley.com/wood/page.aspx?c=1&p=30039&...)

Example: What is the thread count of a no. 6 bolt? http://www08.wolframalpha.com/input/?i=What+is+the+thread+co...

Any engineer would love it.

I used it a lot during the first week, but I'll admit that I'm down to using it maybe once a week. It's just a cumbersome interface that doesn't really make much sense. I'm sure they'll spiffy it up soon.

It was described as a Google Killer by many of the media outlets that covered it. The problem with that description is that it doesn't attempt to do what Google does, so users going to the site hoping to do a search they'd typically send to Google quickly realized it's not for them. Rather than a fad, it will probably continue to be a decent engine that's very useful for a certain segment of users .... but not a viable general-purpose search engine for mainstream users.

More fun with trends

Wolfram, Cuil & bing http://www.google.com/trends?q=wolfram%2C+cuil%2C+bing&c...

Actually need an 18 month range bu this will do.

I'm actually suprised at how much higher Cuil's peak was compared to bing considering the current marketing push (is this all over or is MS pushing harder here in the Bay Area)

i still love it, but it isn't useful for me on on a daily basis.

UI needs to be fixed, here are my thoughts:

http://codingthriller.blogspot.com/2009/05/fixing-wolframalp...

This is why:

http://unqualified-reservations.blogspot.com/2009/07/wolfram...

This is like searching "google" on wolfram alpha trends. Use compete.com if you want to prove a point, not google trends.

It's very useful. I use it daily, particularly for date math.

I have the firefox plugin, but other than that I just use it for math I'm too lazy to figure out myself - or to check my math.

Stupid. Just because traffic is going down doesn't mean the technology or the company doesn't deserve some serious attention:

http://singularityhub.com/2009/07/27/wolfram-alpha-a-force-t...